Silicon Valley has worked through the usual immortality menu: fasting, blood transfusions, freezing your head. This week brings the next tier, which is growing an entire spare you and leaving out the part that thinks. Elsewhere: real movement toward lab-made sperm, a regulator getting rolled by its own advisory panel, and a cancer dataset quietly confirming that the cells we have screened drugs in for decades have been fibbing.
Table of Contents
🧟 The Backup Body Nobody Wants To Admit They're Building
NEWS
Picture a version of yourself that never wakes up. Same DNA, same organs, no cerebral cortex, kept breathing on a feeding tube until the day you need a kidney. That is roughly the pitch MIT Technology Review found R3 Bio's founder John Schloendorn making privately, under the label "brainless clones." One attendee described the briefing as a close encounter of the third kind with Dr. Strangelove.
The biology leans on hydranencephaly, a rare condition in which most of the brain is absent. The logistics are where it gets kinda uncomfortable: artificial wombs do not exist, so the first batch would have to be gestated by paid surrogates. R3 responded with a sweeping denial, and its site now states flatly that the company does not produce the large-scale structures described in the media.
Which sits awkwardly next to cofounder Alice Gilman spending nearly two hours on a podcast describing "organ sacks" before trying to get the episode shelved, calling R3 a "federal asset," and explaining the field's central bind: less nervous system means more social acceptance and less scientific feasibility. She also described the method as doing "some Yamanaka miracles," borrowing the shine of Shinya Yamanaka's Nobel-winning reprogramming work. Nobody at R3 has a Nobel.
Harvard's George Church, who advises in this space, called brainless human bodies not very useful in addition to being repulsive. When your own field's diplomat reaches for "repulsive," the pitch deck may need work.
🔬 Lab-Grown Sperm, Except The Sperm Part Hasn't Happened
RESEARCH
In May we covered a Utah company that announced lab-grown human sperm by press conference, with no paper, no preprint, and no data, and we said we would watch the bioRxiv feed. Someone finally did it properly. Kotaro Sasaki's group at Penn Vet published in Cell Stem Cell, with actual methods attached.
The route is deranged… in the best way. Take blood cells, reprogram them into iPSCs, coax them toward germ cells, then mix them with fetal testicular cells from mice to build a chimeric mini-testis, and incubate the whole thing in a pouch grown on a mouse's kidney. Six months later, the cells had self-organized into tubular structures and become spermatogonia. They also did it with rhesus macaque cells, a first, and the results matched real human and monkey cells by up to 97%.
Now the part the headlines skipped. Spermatogonia are not sperm. They are the precursor cells, several steps upstream. Andrologist Allan Pacey put it plainly: the authors have not made sperm, just early germ cells before they divide and grow tails. Sasaki's own framing is refreshingly unglamorous, putting the field in the middle or a little beyond.
Still a real advance, and the honesty is the tell. The people with the strongest result are making the smallest claim.
💉 The FDA Panel That Outvoted The FDA's Own Scientists
NEWS
The compounded peptide market runs on vibes, testimonials, and vials labeled "research use only." In July, it got closer to a legal supply chain, and the route there is a case study in how rules actually get made.
Start in late 2023, when the FDA parked 19 peptides in Category 2, the bucket for substances posing significant compounding safety risks. Then, in February 2026, HHS Secretary Robert F. Kennedy Jr. announced that roughly 14 of the 19 would move back to Category 1, a policy direction delivered on a podcast rather than a regulatory action. By July, a panel had seven of them in front of it, with almost no high-quality randomized human trials to read.
It voted 8 to 6 with one abstention to recommend BPC-157, KPV and TB-500, and 7 to 5 with two abstentions on MOTS-c, after the agency's own staff reviewers recommended against. Seven of the 14 members had been added ahead of the meeting from clinics or businesses selling peptide treatments. Panelist Elizabeth Rebello of the University of Texas said she was concerned they were responding to market-induced demand rather than solid science.
Worth knowing what is actually being waved through. BPC-157 is a 15-amino-acid fragment originally isolated from human gastric juice, and it has become a gym-bag staple on the strength of tendon and muscle recovery claims.
Which makes the paperwork the best joke in the story. The FDA evaluated BPC-157 for ulcerative colitis, not recovery, and its advisory committees issue non-binding recommendations regardless. Nothing is settled, no final determination has been issued, and a recommendation is not a rule.
🧫 Decades Of Cancer Cell Lines, Politely Debunked
RESEARCH
Cancer biology has a dirty secret it discusses openly and then ignores: the flat cell lines everyone screens drugs in have spent decades in plastic dishes, forgetting what tissue they came from. On August 5, Nature published three papers about it at once, including a 256-organoid biobank from the Sanger Institute and a compendium of patient-derived models. The one we liked most quantified the damage.
A Broad Institute team led by Francisca Vazquez ran 147 genome-scale CRISPR screens across organoids and spheroids from 10 cancer types. Then they asked each model a simple question: can you still tell us which tissue you came from? The 3D models were right 69% of the time, traditional cell lines managed 35%, and for brain tumors the split was 93% versus 11%, with more than a quarter of traditional models having drifted into an undifferentiated state. Decades of "this cell line represents glioblastoma" turns out to have been generous.
The payoff is new drug targets rather than just embarrassment. KRAS-amplified oesophageal organoids proved vulnerable to SCD inhibitors, a dependency invisible in traditional culture because serum is full of the lipids the cells would otherwise have to make themselves. The 2D dishes were feeding the cancer its own answer key.
To their credit, the authors do not oversell. Organoids lost ESR1 dependency entirely, making them useless for the most common breast cancer subtype, and for prostate models the old cell lines actually won. Both types are wrong, just about different things, which is why the combined dataset now sits in one portal rather than replacing one orthodoxy with another.
🚬 A Quarter-Million Public Comments, Almost None From The Public
RESEARCH
When the FDA proposed banning menthol cigarettes and flavored cigars in 2022, the public responded in record numbers, and the volume of that response became a reason to delay. A Rutgers team read all 246,808 comments and traced them back to just 17 distinct form-letter templates.
Their findings in JAMA Network Open: 97% of the menthol comments and 93% of the cigar comments were form letters, and roughly 46% of the menthol form letters traced back to tobacco company campaigns, including Reynolds American's "Own It, Voice It" and Altria's "Voices for Consumer Choice." Just 2% of all comments were unique.
The machinery is worth describing, and it comes from a separate Rutgers paper published last year. Companies emailed customers urging them to take action and steered them to company-run sites where you first picked which sort of person you were, adult menthol consumer, business owner, concerned citizen, or farmer, then picked which anti-ban arguments worried you most. The site assembled the letter and filed it. Every submission looked like a citizen. Collectively, they looked like a groundswell.
Lead author Andrea Villanti noted that the sheer volume was used to justify delaying a policy that would have saved lives. The proposed rules were withdrawn in 2025.
Nothing here was illegal. Form comments are standard practice across every industry and advocacy group in Washington. Which is the uncomfortable part: the system worked exactly as designed, and the design is the problem.
So what could we see this week? It is the distance between what the data says and what the story says. A startup describes spare human bodies in private and denies it in public. A careful paper about precursor cells becomes "lab-grown sperm" by the time it reaches your feed. A panel outvotes its own scientists. Decades of cell lines turn out to have been quietly drifting. And a quarter-million voices turn out to be a mail-merge with a tobacco logo on it.
The comforting part is that every one of those gaps was closed by someone doing the boring work: reading all 246,808 comments, running 147 CRISPR screens, checking whether the cells actually became sperm. Nobody live-tweets that part.
Which of these landed hardest? Hit reply; we read everything. And if you know someone who would enjoy learning that their backup body is behind schedule, forward this along.
Keep questioning everything (especially anything described as categorically false),
P.S. We ran this edition past our legal team and they confirmed the newsletter reserves the right to hold hypothetical futuristic discussions.